Information Quality Assurance by Lazy Exploration of Information Source Combinations Space in Open Multi-Agent Systems

Jisun Park (The Laboratory for Intelligent Processes and Systems The University of Texas, USA)

K. Suzanne Barber (The Laboratory for Intelligent Processes and Systems The University of Texas, USA)

Abstract: Information quality assurance under the existence of uncertainty can be investigated in the context of soft security, where an agent maintains trustworthiness evaluations of its information sources to assist in the evaluation of incoming information quality from those sources. Since dependency inherently exists in a system where agents do not have self-sufficient sensing or data collection capabilities, finding an appropriate set of information sources is important for assuring the quality of information and for increasing the agent's goal achievement. This research proposes an approach for selecting information sources as partners. In order to increase the efficiency and the accuracy, we use trustworthiness, information cost and goal coverage as the metrics for information valuation while adopting a lazy exploration of information sources combination space. Experimental results show that the proposed approach increases the efficiency and results in quality information acquisition.

Keywords: agent, information quality, multi-agent systems

Categories: H.3.3, I.2.11, I.2.8